Next Generation Multiscale Adaptive Mesh Atmospheric ...Navier-Stokes Equations: Full discretised...

Next Generation Multiscale Adaptive

Mesh Atmospheric Modelling, Rapid

Response and Data Assimilation

Fangxin Fang, Jeff Gomes

29, August, 2017

AMCG website (http://www.imperial.ac.uk/earth-science/research/research-

groups/amcg/)

Fluidity❖Open Source Model Software for

Multiphysics Problems

❖Unstructured FEM Meshes

❖Anisotropic Adaptive Mesh

technology

❖User-friendly GUI

❖Python interface to calculate

diagnostic fields, to set

prescribed fields and user-

defined boundary conditions

Atmospheric Modelling

Ocean, flooding multifluid flow

modelling

Data science laboratory

HPC

Uncertainty quantification

➢ Analysis, real-time updating and

representation of uncertainties.

Reduced order modelling

➢ Emergency rapid response

Data assimilation➢ Optimisation of

uncertainties➢ Minimisation of misfit

between observational data and numerical results

Adaptive targeted observations➢ Optimal

experimental design

➢ Optimal sensor locations

Multi-scale predicative modelling with multi-physical Fluidity

Optimal design, risk assessment, decision and policy-making processes risk

Predictive and Uncertainty Model Framework

Computational Multi-Fluid Flow Dynamic Model❖ Porous media model embedded in Fluidity are based on:

• new family of FE element-pairs (PnDG-Pm and PnDG-PmDG) and

• numerical formulation (overlapping CVFEM) that ensures high-order

accuracy on the solution fields (i.e., pressure, velocity, saturation,

temperature etc)

❖ The model has been used on transport of contaminant in subsurface media

(mining spillage), oil and gas production, nuclear waste repository etc;

Atmospheric Environmental Model: Critical bridge between

human activities and environmental change (IAP)

Clean Day PM2.5~15 µg m-3

Polluted Day PM2.5~200 µg m-3Severely Polluted Day PM2.5~470 µg m-3

Beijing, Olympic site

Human Activities(urbanization, motorization and

industrialization, emissions)

CO2, CO, NO2, SO2, O3, aerosols, etc

Atmospheric dynamical transport

Atmospheric chemical transformation

Ocean System

Land , Water, Snows

Terrestrial Ecosystems

Human Health Effects

Agriculture and Ecosystem

Regional air pollution

Global air pollution

Deposition (Acid rain, etc)Natural emissions(carbon,

nitrogen)

Climate Change

Radiative forcing

(scatting ,absorbing, cloud)

Climate issues

Environmental issues

Atmospheric Environmental Model: Critical bridge between

human activities and environmental change (IAP)

Air pollution is largely

influenced by human

activities, and can be

reduced by human.

Urban Air Pollution Modelling

(MAGIC, supported by EPSRC) @Elephant Castle, London

Air Pollutant Modelling

(IAP-ICL, supported by NSFC/ EPSRC)

SO2 released

from over 100

power plants

(over Beijing

and 55 cities)

Chemical modelling (IAP-ICL, Zheng etc.) (NO and NO2 released from over 100 power plants)

Continuous formation/consumption of NOx and Ozone over 5 days.

3D chemical modelling (IAP-ICL, Zheng etc.) (NO and NO2 released from over 100 power plants)

1 day

2010-03-19_04 2010-03-20_04

2010-03-20_18 2010-03-21_04

Dust storm

(IAP-ICL,

Zheng etc.)

CloudAdaptive mesh

3D Atmospheric Modelling: Cyclone 3D Simulation (IAP-ICL)

Rain

Potential Temperature

Velocity Vector

Vapor Water

Tohoku event from 2011

(Visiting PhD student from Japan)

Fig. 11. Situation of Study Area in Greve, Municipality of Denmark, see Soledad (2014)

Natural Disaster: Simulating Flooding – Denmark

(R. Hu etc. supported by the EU PEARL project)

Surface Surface with mesh

Natural Disaster: Simulating Flooding – Denmark

(R. Hu etc. supported by the EU PEARL project)

Atmospheric Modelling

Ocean, flooding multifluid flow

modelling

Data science laboratory HPC

Uncertainty quantification

➢ Analysis, real-time updating and

representation of uncertainties.

Reduced order modelling

➢ Emergency rapid response

Data assimilation➢ Optimisation of

uncertainties➢ Minimisation of misfit

between observational data and numerical results

Adaptive targeted observations➢ Optimal

experimental design

➢ Optimal sensor locations

Multi-scale predicative modelling with multi-physical Fluidity

Optimal design, risk assessment, decision and policy-making processes risk

Predictive and Uncertainty Model Framework

Data Assimilation (DA)DA methods:

❖ Optimal interpolation;

❖ Nudging;

❖ 3D-Var;

❖ 4D-Var (Adjoint);

❖ Ensemble KF

Motivation for DA:

❖ To improve the predictability of numerical

models;

❖ Uncertainty sensitivity analysis;

❖ Optimisation of uncertainties in models;

❖ Goal-based error measure and mesh

adaptivity;

❖ Design optimisation;

❖ Adaptive observation (Optimisation of

sensors locations).

PM2.5 before DA

Differences

PM2.5 after DA

Sources: CNEMC

IAP

Work from Institute of Atmospheric and Physics, China

DA: Optimal sensor location

Navier-Stokes Equations:

Full discretised system (high dimensional):

Projecting onto the reduced space using SVD/POD

Discretising equations (1) and (2)

Wring the reduced order model in a general form

The function f is

represented

by deep learning

Reduced Order Modelling

Reduced Order Modelling (POD) and Deep Learning (Xiao etc)

Top panel: Full modelling

Rapid modelling: Air flow (Elephant Castle London)

(Xiao etc. supported by MAGIC –EPSRC)

Left: reduced order modelling; right: high fidelity modelling

CPU time: seconds (Reduced order model); 3 hours (10 cores, Full fidelity model)

Future work

HPCDA

Fluidity

NAQPM WRF

Adaptive-mesh air quality model

Gas chemistry

Adaptive-mesh weather forecast

Liquid chemistry

PM chemistry

Microphysics

Cumulus parameterization

Surface Layer

Planetary Boundary Layer, Radiation

Reginal Model Global Model

Future workFuture work

Thanks

Rapid modelling: Flow past two buildingsTop: reduced order modelling; bottom: high fidelity

modelling